College and Research Libraries Organizational Commitment of Professional Employees in Union and Nonunion Research Libraries Tina Maragou Hovekamp This study was a comparative exploration of organizational (library) commit- ment among professional library employees in unionized and nonunionized research libraries. Based on a survey of 400 individuals, the study examined organizational commitment and its relationship to (1) union presence; (2) union membership status; and (3) union commitment. The results of the statistical analysis revealed that union presence was a negative predictor of organizational loyalty; registered and nonregistered union members shared similar attitudes toward their employing organization; and union commitment tended to be a positive predictor of organizational commitment. The conflicting findings in the test of the relationships between organizational commitment and union presence and between organizational commitment and union com- mitment might be partially attributed to the fact that union commitment was relatively low among the library professionals. In addition, such results suggest that among union libraries factors other than union loyalty may negatively affect employee feelings toward their library. • nion organization has often generated concerns about the possible conflicts it might in- troduce in the relationship be- tween employees and employer. Such concerns have also been expressed in the library literature, however, without any attempt to investigate in a systematic way the consequences of unionization for the profession's allegiances. Even outside the library field the topic of un- ion impact on the employees' loyalty to their employer has received only sporadic research attention, mainly in the area of blue-collar workers. However, research findings on this subject can have a particu- lar importance among professional em- ployee~in this group of workers the issue of unionization and its potential consequences have been controversial for at least the last two decades. This study was designed to investi- gate any associations between unions and commitment to the library as an employer. It was based on survey results from library employees who had an M.L.S. or equivalent degree and held positions with professional classifica- tion (academic or faculty status) in unionized and nonunionized research libraries. The data received from the two groups, union and nonunion, were compared in order to examine the r.ela- tionship between commitment to the library as an employing organization and (1) union presence; (2) union Tina Mamgou Hovekamp is Libmry Instruction and Public Seroices Librarian at the Leslie J. Samge Library, Western State College of Colorado, Gunnison, Colorado81231. This article is lxzsed on tile author's dissertation at the University of North Carolina at Chapel Hill. This research was supported by a CLR/Kellogg dissertation gmnt. 297 ~--------------------------------------------------------------~-------- 298 College & Research Libraries membership status; and (3) union com- mitment. Multiple regression analysis was used with the control of several demographic variables that could otherwise affect the results of the statistical analysis. Al- though these control variables were not of central concern in the research, their entrance into the regression equations al- lowed the examination of their possible impact on the way the surveyed employ- ees felt about their library employer. BACKGROUND OF THE STUDY The emergence of unions can be treated as the result of conflict of inter- ests between management and the workers. Ralf Dahrendorf suggests that such conflicting goals always exist in all authority relations, but it takes the con- scious organization of opposing groups to make them clear and articulate to their members.1 Dahrendorf' s basic argument has been that social organizations, and in fact all of society, are characterized by authority relations. Authority exists "wherever there are people whose actions are subject to legitimate and sanctioned prescriptions that originate outside them but within the social structure."2 The implication is that authority, in any form, is an inherent element in the social structure, independent of the personal- ity of the people who exercise it. One of the most important proposi- tions of Dahrendorf' s theory is that authority relations are the cause of po- tential clashes of interests between those endowed with the decision-making power and those who are subject to it. Under certain conditions, Dahrendorf says, these clashes generate the forma- tion of interest groups that attempt to modify the characteristics of the social structure. The purpose of such groups is to improve the status of their members and increase the probability of finding satisfaction for their wants and needs. Thus, group conflict can be perceived as an inescapable consequence of the presence of opposing interests, and as Lewis A. Coser defines it, "a struggle over values and claims to scarce status, July1994 power and resources."3 It is true that conflict often carries with it negative connotations that come from the very fact that it is usually associated with ag- gression and hostilitY, Yet, as Coser ex- plains, "aggression may be regarded as an index of conflict, but this does not imply that every conflict must be accom- panied by aggressiveness."4 A number of sociologists believe that the existence of conflict is not necessarily detrimental or dysfunctional for the relationships in which it occurs. Some prefer to treat it as a balancing force in power distribution and, for this reason, as valuable and bene- ficial. Georg Simmel, for example, de- scribed conflict in the following terms: Conflict is designed to resolve di- vergent dualisms; it is a way of achiev- ing some kind of unity. . . . This is roughly parallel to the fact that it is a most violent symptom of a disease which represents the effort of the or- ganism to free itself of disturbances and damages caused by them. . .. Conflict itself resolves the tension be- tween contrasts. 5 In this way, Simmel advances the no- tion that conflict serves as a "safety valve" for the release of hostility which otherwise would be detrimental in the relation between the antagonists. By re- leasing feelings of resentment, conflicts can work as a force for maintaining a relationship. Employee organization in the form of either unions or other independent em- ployee associations is one example of contemporary experience of social con- flicts. The beginning of the twentieth century saw the phenomenal rise of big business and big government. As C. Wright Mills describes, both in the pri- vate industry and public sector, employees often found themselves in confrontation with complex and depersonalized em- ployment relations. In such a bureauc- racy, unions appeared to some of these employees as one way to exert some in- fluence in the creation of work rules.6 However, by the 1950s the fast expan- sion of unionization brought with it deep concern about the organized em- ployees' degree of commitment to their employing organization. According to Lois Dean, a number of people consid- ered it a matter of common sense that workers who belonged to or supported union organizations perceived a funda- mental conflict of interest with manage- ment, which caused them to identify less with the employing organization.7 Scholars of that period, for instance, Solomon Barkin and George W. Eng- land, argued that workers cannot main- tain loyalty to both their unions and companies at the same time.8•9 England tried to prove his thesis with the evi- dence of a study he conducted within two unionized employee groups. His re- search, however, showed that there was actually no strong relationship between attitudes toward the company and atti- tudes toward the union. These suggestions gave rise to re- search interest and were soon chal- lenged by new research findings which indicated that rank and file loyalty to the union did not necessarily preclude loy- alty to management. 10 In 1952, Arnold Rose, in a study of a Teamsters' local in St. Louis, said, "People can have loyalty to two or more groups or two sets of values, even when those groups or val- ues are in conflict. In concrete terms, loyalty to the union does not mean dis- loyalty to the employer." 11 Two years later, Dean found that posi- tive attitudes toward both union and management may occur regardless of the degree of conflict between union and management. More specifically, using data from three union studies, Dean re- ported that even in cases of overt conflict between the two organizations, the worker's dual loyalty may still exist in the plant. Also in this study, the workers who were found to feel most positively toward management were equally well- disposed towards the union, and vice versa.U Dean's research findings were confirmed by Ross Stagner in his lllini City investigation, and Theodore Purcell in his research on the concept of dual allegiance in a meat packing plant in Chicago.13•14 Both of these studies showed that dual commitment is possi- ble among union members. Organizational Commitment 299 In spite of all the research ferment that the concept of dual allegiance created in the 1950s, the subject lay dormant for nearly twenty years. In the meantime, the research interest shifted to the concept of organizational commitment, which, ac- cording to Harold L. Angle and James L. Perry, later inspired the resurgence of the theme of dual commitment.15 In spite of all the research ferment that the concept of dual allegiance created in the 1950s, the subject lay dormant for nearly twenty years. In recent years, a number of studies have tried to explain variances in the degree of dual allegiance among rank and file union members. For example, Cynthia Fukami and Erik Larson exam- ined the degree of expressed dual com- mitment among unionized blue-collar workers in the private sector.16 Their analysis of data initially indicated a sig- nificant correlation between union and organizational commitment. However, when the members' perceptions of the nature of the union-management rela- tions were controlled, this relationship was insignificant. Fukami and Larson pointed out that union employees may be equally committed to their employer and union, but that the strength of this dual commitment depends on the way workers experience union-management relations in their workplace. Taking a similar approach, Angle and Perry investigated twenty-two munici- pal unionized bus companies. These re- searchers verified that the extent of expressed dual commitment is related to the nature of the labor-management re- lationship and the degree of cooperation between the two. In addition, Angle and Perry found that dual allegiance was moderated also by the worker's degree of union participation. The more active members were, the weaker the relation- ship between organizational and union commitment. The researchers offered no theoretical explanation in relation to this finding. 17 300 College & Research Libraries Another recent study focusing on the same issue was James Martin's examina- tion of the concept of dual allegiance in a federal government facility in a large Midwestern city.18 In contrast to pre- vious studies, Martin analyzed re- sponses of · both union and ·nonunion members within a single bargaining unit. An interesting result was that the largest portion of union meml>ers expressed posi- tive attitudes toward both union and man- agement. Conversely, the majority of nonmeml>ers had negative attitudes to- ward both union and management. Mar- tin tried to explain his findings as a result of the union's cooperative relations with management. Unfortunately, Martin's conclusions suffer from limited applica- bility because of the very small sample of participants he used. Daniel Gallagher was the second to attempt an examination of the relation- ship of union versus organization com- mitment on a similar sample of federal government employees.19 Like Martin, Gallagher tried to compare the relation- ship of union and organizational commit- ment l>etween dues-paying meml>ers and nonmeml>ers who l>elonged in the same bargaining unit. Gallagher found that the levels of commitment to the employer for both union meml>ers and those who had never l>een members were equivalent. In contrast to Martin, Gallagher indicated that nonmeml>ers did not have negative feelings toward their employer but they reported higher commitment to the em- ploying organization rather than to the union. In his conclusion Gallagher also argued that "a worker's decision to join a union should not be interpreted as evi- dence of low commitment to the organi- zation."20 Gallagher's statement parallels the conflict theorists' belief that conflict is a natural, and not necessarily a disrup- tive, process. In general, as Angle and Perry indi- cate, the above studies show that "dual commitment is a relatively common phenomenon." 21 Along these lines the present study tried to verify the degree of organizational commitment, this time among professional union employees. Union organizing and its relation to cer- July 1994 tain work attitudes, especially among professionals, has attracted many argu- ments but rather limited research atten- tion. Even those few investigations which dealt with unioris and their effect on the professional employees' attitudes mainly focused on those factors influ-. encing voting l>ehavior during union elections. Researchers have not yet paid much attention to how these employee organizations might affect employee at- titudes after the organizations are estab- lished. The scarcity of such research is particularly evident in the professional library literature where the question of conflict between unionized library em- ployees and management has generated many arguments, yet no empirical evi- dence to support them. Consequently the present study was designed to· ad- dress the relationship between union presence and commitment to the em- ploying organization among profes- sional employees in libraries. For the purposes of this investigation, Richard Mowday, Lynman Porter, and Richard Steer's definition of organiza- tional commitment was employed. Or- ganizational commitment is the result of an employee's having: a. a strong belief in and acceptance of the organization's goals and values; b. a willingness to exert considerable effort on l>ehalf of the organization; and c. a strong desire to maintain mem- bership in the organization.22 It should be noted that the terms or- ganizational commitment and library com- mitment are used interchangeably in this . study to mean the way the respondents feel about their library as an employer. The questions of the survey instrument measured the participants' sentiments about their library as an employing or- ganization. HYPOTHESES The hypotheses of this study tested the possibility of differences in organiza- tional commitment among professional librarians in union versus nonunion re- search libraries and registered versus nonregistered union members. The rela- tionship between union commitment and organizational commitment was also investigated. The questions tested included: 1. Is there a significant relationship between the presence of unions and professional librarians' com- mitment to their library? 2. Is there a significant relationship between union membership status and professional librarians' com- mitment to their library? 3. Is there a significant relationship between the professional librari- ans' commitment to the union and their commitment to their library? METHOD Subjects and Setting13 The subjects of this study were em- ployees with professional appointments in academic research libraries that are members of the American Research Li- brary Association (ARL). For the purpose of the present research, professional li- brary employees were defined as full-time or part-time library employees who had an M.L.S. or equivalent degree and held positions with professional classifica- tion (academic or faculty status). In August 1991 requests for participa- tion were mailed to twenty-six United States research libraries which at the time did not have union representation for either professional or paraprofessional staff. These libraries were selected by eliminating from a list of ARL academic libraries those which had bargaining agreements for professional and/ or para- professionallibrary staff. The sample was then drawn from the remaining popula- tion, using a table of random numbers. At the same time another mailing was sent to all seventeen American research libraries that had union representation for their professional library employees. The reason for including a larger sample of nonunion libraries was that the major- ity of ARL institutions did not at the time have bargaining agreements. In order to achieve a more accurate representation, the researcher decided to include a larger number of libraries without union representation. Organizational Commitment 301 The letter to both types of institutions was addressed to library directors and explained the purpose of the study and asked for their participation. Nineteen nonunion research libraries (i.e., 73 per- cent of the original nonunion library sample) and thirteen union research libraries (i.e., 76 percent of the union library sample) agreed to participate. Based on lists of employee names pro- vided by the libraries, the researcher cal- culated a percentage of individuals that would be asked to participate in the study so that the total number of partici- pants for each of the two groups, union and nonunion, was two hundred. Ex- cluded from the population were law and health library professionals, as well as librarians at regional campuses. The reason for this exclusion was that among campuses there was no consistency in whether law, health, or regional campus libraries were part of the central research library system or whether they operated independently from it. In an effort to create a more homogeneous population profile, the researcher decided to ex- clude them from the survey. Also ex- cluded from the nonunion list were higher administrative staff, that is, direc- tors, assistant directors, and personnel administrators, because this type of em- ployee was not part of the union group (higher administrative staff are not rep- resented by unions). Finally, the names of participants in the survey were se- lected using a table of random numbers. Measures Commitment to the employing or- ganization was measured by Porter's Organizational Commitment Question- naire (OCQ).24 Minor modifications were made in the language of this ques- tionnaire to make it more suitable to the surveyed population. OCQ has been used extensively in previous studies that measured worker commitment to em- ployers. Previous research has shown sufficient levels of reliability and valid- ity as a general measure of commitment for most working populations. In addi- tion, the definition of organizational commitment used in the construction of 302 College & Research Libraries OCQ was consistent with the purposes of this investigation. The union commitment measure of this study included eleven items adapted from Porter's Organizational Commitment Questionnaire (OCQ). A very similar instrument was previously used in union studies by Chester A. Schriesheim and Anne S. Tsui, Dan R. Dalton and William D. Todor, and Ed- ward J. Conlon and Daniel J. Gal- lagher.2s--27 A few modifications in the language of this measure were included to better fit the purposes of this study. This questionnaire had shown high lev- els of internal reliability and agreed with the definition of union commitment in the present study. The items of this measure were included only in the sur- vey of the union participants. The response format for the items of both OCQ and union commitment meas- ure employed a 5-point Likert Scale (1 to 5) with anchors ranging from "strongly agree" to "strongly disagree." Participants in the union sample were asked to identify themselves as regis- tered or nonregistered union members. Based on the design of similar past stud- ies, all respondents were asked to indi- cate their gender, age, total years of work as a library employee, full-time or part- time employment status, and total an- nual salary. In the present research these demographic variables entered the sta- tistical analysis as control variables. In this way it was also possible to explore the extent to which they may also have an effect on organizational commitment. The last page of the questionnaire in- vited the participants to offer additional comments or other information that they thought might be valuable to the study. These comments were used to speculate further on the results of this research. Data were collected during fall 1991. Out of the total sample of 400 surveyed individuals, 363 or 91 percent re- sponded, including six refusals to par- ticipate and thirteen invalid responses. Excluding the refusals and invalid re- sponses, the response rate reached 89 percent. Within the nonunion group July 1994 there were 189 or 94.5 percent returned responses, whereas within the union group there were 174 or 87 percent re- turned responses. RESULTS Table 1 provides the population pro- file of this survey based on an analysis of the demographic characteristics of the respondents. It should be noted that the distribution of the population regarding employment status (full-time versus part-time) was quite uneven. However, the researcher de- cided to include this demographic vari- able in the analysis of data. TABLEl SUMMARY OF DEMOGRAPHIC CHARACTERISTICS FOR RESPONDENTS Demographic %of Characteristics Res2ondents Gender Male 36.3 Female 63.7 Age 20-30 5.3 31-40 29.6 41-50 41.6 51-60 15.0 over 61 8.5 Years of library experience 1-5 12.8 6-10 18.1 11-15 18.7 16-20 20.2 21-25 15.8 26-30 7.0 over 31 7.4 Employment status Full-time 93.9 Part-time 6.1 Annual salary less than $20,000 3.2 20,000 to 24,999 7.3 25,000 to 29,999 19.8 30,000 to 34,999 17.5 35,000 to 39,999 17.3 40,000 to 44,999 17.3 over $45,000 17.6 Organizational Commitment 303 TABLE2 REGRESSION RESULTS PREDICTING LIBRARY COMMITMENT BY TYPE OF INSTITUTION (UNION VERSUS NONUNION) WITH DEMOGRAPHICS AS CONTROL VARIABLES Independent Variables Coefficient T SignT Union -4.860 -3.998 0.000*** Gender 2.255 1.838 0.067 Age 0.239 0.278 0.781 Years of library experience -0.345 -0.628 0.530 Part-time (versus full-time) 6.160 2.222 0.027* Salary 1.553 3.382 0.001** Constant 42.674 Note: The whole regression model had an R2 = 0.076, with F = 4.550, p < .001 (df = 6,331). N = 344; • p < .OS; •• p < .01; ... p < .001 The means of organizational commit- ment among professionals in unionized and nonunionized institutions were quite moderate, 3.114 and 3.375 respectively, in- dicating that, in general, the surveyed population seemed to feel more or less neutral about their library employer. Multiple regression analysis was used for the test of all three hypotheses of the study. This procedure allowed for the control of demographic variables such as gender, age, years of library experi- ence, employment status (full-time ver- sus part- time), and salary, which were included in all regression equations. Table 2 presents the findings of the regression analysis for the first hypothe- sis of the study which tested the signifi- cance of the relationship between union presence and professional librarians' commitment to their library. The results indicated that the presence of unions, part-time employment status, and salary were statistically significant contributing factors to overall library commitment. Specifically, union pres- ence was negatively related to library commitment, while the two demographic variables of part-time employment and salary had a significant positive relation with library commitment. Professional employees in unionized libraries re- ported lower feelings of commitment to their library. In addition, those with part-time appointments and higher sala- ries tended to identify more strongly with their employer. The above regression model had a relatively low R2 value. Only 7.6 percent of the variance of overall library com- mitment could be explained by the variables of union presence and the demographic control variables (particu- larly, part-time status and salary). This study investigated the issue of organizational commitment within the union group. Specifically, it explored the possibility of a difference in the feelings of organizational loyalty among profes- sional library employees who were reg- istered union members and those who were not registered members but were covered by their library's collective bar- gaining agreement. Table 3 presents the results of the regression analysis. Union membership, after controlling for demographic characteristics, had no significant relationship with organiza- tional commitment. Salary, once again, proved to have a positive relationship with commitment to the library. As a matter of fact, 11.6 percent of the vari- ance of library commitment among the respondents in unionized institutions was explained primarily by the salary variable. Those participants with high salaries tended to report stronger com- mitment to their library regardless of whether they were registered or nonreg- istered union members. 304 College & Research Libraries July 1994 TABLE3 REGRESSION RESULTS PREDICTING COMMITMENT TO THE LIBRARY BY TYPE OF UNION MEMBERSHIP (REGISTERED VERSUS NONREGISTERED MEMBERS) WITH DEMOGRAPHICS AS CONTROL VARIABLES Coefficient T Sign T Union membership 1.582 0.794 0.429 Gender 3.664 1.903 0.059 Age 1.090 0.824 0.411 Years of library experience -1.482 -1.838 0.068 Part-time (versus full-time) 6.387 1.524 0.130 Salary 2.538 3.693 O.OOQ>t•• Constant 32.511 Note: The whole regression model had an R2 = 0.116, with F = 3.205, p < .01 (df = 6,146). N = 155; ••• p < .001 TABLE4 REGRESSION RESULTS PREDICTING THE RELATION OF LIBRARY COMMITMENT TO UNION COMMITMENT WITH DEMOGRAPHICS AS CONTROL VARIABLES Independent Variable Coefficient T SignT Union commitment 0.232 2.482 1.971 0.014 .. 0.051 0.661 0.110 0.075 0.()()()'1' .... Gender 3.718 Age 0.579 0.440 Years of library experience -1.279 -1.608 Part-time (versus full-time) 7.317 1.791 Salary 2.622 3.931 Constant 26.186 Note: The whole regression model had an R2 = 0.148, with F = 4.244, p < .01